March 02, 2017Douglas Booms
Voice technology carries enormous promise, as many people know from their interactions with Alexa, the voice service that powers devices like Amazon Echo. Our teams here at Amazon are working hard to unlock that potential, but innovation requires a collective effort, from large companies like ours to two-person startups, and from casual hobbyists to major academic institutions.
Today, we’re excited to announce the Alexa Fund Fellowship, a new program to support universities and researchers focused on transformative voice technologies such as text-to-speech (TTS), natural language understanding (NLU), automatic speech recognition (ASR) and conversational artificial intelligence (AI).
Alexa Fund Fellows will receive funding, access to Alexa devices, and mentoring from an Alexa Science team member to develop an undergraduate or graduate curriculum around one or more of these disciplines. The year-long program culminates in a demo day where students can showcase their work for peers, university faculty and members of the Alexa team.
We’re also pleased to announce our first group of university participants: Carnegie Mellon University (CMU), Johns Hopkins University, University of Southern California (USC), and University of Waterloo. Both Carnegie Mellon and Waterloo have already chosen fellows and begun offering classes for engineering students, while courses from Johns Hopkins and USC will begin in fall 2017.
Carnegie Mellon’s Alexa Fund Fellow, Ran Zhao, is assisting this spring in teaching “Dialog Systems.” Course participants will learn how to implement a complete spoken language system while exploring research topics of interest in the context of a functioning system. The course will emphasize issues in task design and dialog management and in language generation and synthesis.
Alexa Fund Fellow Chahid Ouali will support technology integration between students and Amazon while supporting faculty with courses such as, “Fundamentals of Computational Intelligence,” which will introduce students to novel approaches for computational intelligence based techniques including knowledge-based reasoning, expert systems, fuzzy inferencing and connectionist modeling based on artificial neural networks. The focus is on designing computationally intelligent systems with human-like capabilities in reasoning, learning, and adaptation.
“Voice computing is no longer science fiction. By understanding the principles behind the interaction between computers and humans, our students can develop new applications and start-up ventures,” said Pearl Sullivan, dean of the Faculty of Engineering at University of Waterloo. “With our intense focus on experiential learning and early innovation, this program with Amazon will enable our engineering faculty to use Alexa as a teaching tool for artificial intelligence topics on voice recognition and speech synthesis.”
Professor Fakhri Karray works with students to set up Amazon Echo devices in the University of Waterloo Engineering lab. Photo provided by the University of Waterloo.
At Johns Hopkins, the Alexa Fund Fellowship will support a doctoral student, called the Alexa Fund Fellow, who will mentor and provide technical advice to students enrolled in a master’s degree program that focuses on human language technology, including speech recognition and text understanding. In this newly created program, groups of two to four master’s students will team up to devise and test novel algorithms for speech and natural language processing and speaker recognition. The Alexa Fellow will offer guidance as these team projects evolve over two semesters.
“We are pleased to deepen our relationship with Amazon through the Alexa Fund’s support of our PhD program in our Center for Language and Speech Processing,” said Ed Schlesinger, the Benjamin T. Rome Dean of the Whiting School of Engineering at The Johns Hopkins University. “Advancing our relationships with corporate leaders such as Amazon will assist us in our mission to continue to make positive impacts on society through our research, education, and translational activities.”
Alexa will be integrated into the academic and innovation ecosystem at the USC Viterbi School of Engineering and its Information Sciences Institute (ISI). It will serve as a project infrastructure for undergraduate and graduate-level courses in artificial intelligence and intelligent agents. In addition, by studying Alexa, students may develop applications which result in new ventures incubated by the USC Viterbi Startup Garage at ISI in collaboration with the Alexa Fund.
"Alexa is a highly visible driver of the global cognitive AI movement. Our program with the Amazon Alexa Fund opens up exciting and timely new opportunities for our students, for the USC Information Sciences Institute, and for the Viterbi School of Engineering,” said Prem Natarajan, the Michael Keston Executive Director of USC Information Sciences Institute at USC.
We’ve barely scratched the surface of what’s possible in artificial intelligence, machine learning, and speech science, and we believe more education and research among academic institutions is key to fueling progress in the space. We also recognize the potential for this research to produce viable business concepts. The initial cohort of universities was selected in part for the quality of their technology transfer and entrepreneurship programs, and we’re excited to work with them to support Alexa Fund Fellows interested in commercializing their work.
The Alexa Fund Fellowship is currently being run in private beta. Universities interested in learning more about the program can find additional information on the Alexa Fund Fellowship website.
Doug Booms, VP of Worldwide Corporate Development, Amazon